These are all great comments. I'll throw in 3 points.
I would reiterate the point of being selective about what your robot can do. We have made conscious decisions to not add features other teams have thought were critical, whether it was HP load/cycle (2013), high goal shooting (2014) or building 6-stacks (2015). Nor did we build a high climber in 2013. We had successful years in each.
Which leads me to the second point--focusing on how you can work well within an alliance and how you might manage an alliance. This is a team game, not individual. You should consider your overall role from the get go, not when you get to the competition site. That was reflected in the choices we made in each of the years I listed.
And third, building and operating a reliable robot is always a plus. Focus on build quality and use rigorous check lists to ensure that you're always performing well. Alliance captains will notice.
We also have workshop videos on designing and building robots at
www.citruscircuits.org.